CG-AL-X278
Description
Per-model results
| Model | Attempt 1 | Attempt 2 | Avg score | Tokens / attempt | Runs |
|---|---|---|---|---|---|
| Claude Opus 5 | ✗ | ✓ | 81.3 / 100 | 12k / 17k | 3 |
| Claude Opus 5.5 | ✗ | ✓ | 81.3 / 100 | 12k / 4.7k | 3 |
| GPT-5.6 Sol | ✗ | ✓ | 81.3 / 100 | 7.6k / 4.8k | 3 |
| GPT-6 Sol | ✗ | ✓ | 81.3 / 100 | 7.1k / 2.9k | 3 |
| Claude Sonnet 5 | ✗ | ✓ | 75.0 / 100 | 13k / 17k | 3 |
| GPT-6.1 Sol | ✗ | ✓ | 75.0 / 100 | 6.7k / 2.8k | 3 |
| GPT-5.6 Terra | ✗ | ✓ | 68.8 / 100 | 7.5k / 3.4k | 3 |
| GPT-6 Luna | ✗ | ✓ | 68.8 / 100 | 6.8k / 4.4k | 3 |
| Gemini 3.8 Flash (OpenRouter) | ✗ | ✓ | 68.8 / 100 | 9.1k / 35k | 3 |
| Claude Haiku 4 5 | ✗ | ✗ | 62.5 / 100 | 10k / 1.9k | 3 |
| GPT-5.6 Luna | ✗ | ✓ | 58.3 / 100 | 7.3k / 3.2k | 3 |
| MiniMax M3 | ✗ | ✓ | 58.3 / 100 | 7.4k / 17k | 3 |
| GLM 5.3 (OpenRouter) | ✗ | ✗ | 31.3 / 100 | 7.3k / 59k | 2 |